There is increasing demand for renewable energy using photovoltaic technology. In particular, photovoltaic cells are commonly formed on crystalline silicon wafers which are conventionally obtained by slicing a silicon ingot. This process, which typically yields a silicon wafer thicker than 115 μm, wastes a substantial amount of silicon by consuming up to 50% of the silicon body in kerf loss. The resulting wafers are also much thicker than is needed for useful photovoltaic devices.
Thinner silicon laminae have been made by exfoliation of film by heating after high dose ion implantation, typically with H+ ions. However, to make useful silicon laminae by exfoliation for photovoltaic applications, it is necessary to implant ions at high energy, in order to create a weakness layer at sufficient depth.
Also, in order to provide relatively high productivity, it is desirable to employ high beam currents. Implant beams with an ionic current of 100 mA, and energies of 1 MeV, are now being contemplated. The effective beam power delivered to substrates being implanted can be in the order of 100 kW or higher. The need to prevent the substrates being heated by such high implant beam power to excessive temperatures presents a considerable challenge.
In a known type of ion implantation tool, a beam of ions to be implanted is directed at substrates to be implanted (typically silicon wafers) mounted in a batch around the periphery of a process wheel. The process wheel or rotary scan assembly is mounted for rotation about an axis so that the wafers on the wheel pass one after the other through the ion beam. In this way, the power of the ion beam can be shared between the wafers in the batch on the process wheel. The wafers are mounted on substrate holders on the process wheel. The substrate holders comprise a heat sinking surface for supporting the wafer. Forced cooling of the heat sinking surfaces is typically provided by means of water cooling structures.
Contact between the wafers and the heat sinking support surfaces is maintained by canting the support surfaces inwards towards the axis of rotation, whereby the wafers are pressed by centrifugal force against the support surfaces as the process wheel rotates.
The effectiveness of the cooling of the wafers in such implant apparatus using a rotary scan assembly can be dependant on the force with which wafers are pressed against the underlying heat sink surface. There are known ion implant apparatuses which provide a rotary scan assembly in the form of a drum, with the wafers mounted around the interior face of the drum, substantially facing the axis of rotation. This arrangement maximizes the effect of centrifugal force on the wafers to optimize wafer cooling during the implant process.
However, such rotary drum type ion implant apparatuses have not found favour for use when ions are to be implanted to a precise depth in the substrate. This is because the ion beam is directed at the wafers mounted around the interior of the rotary drum in a final beam direction which is at a substantial angle to the axis of rotation of the drum assembly, typically at an angle close to 90°. The resulting variation in the angle between the wafer being implanted and the implant beam, as each wafer traverses the beam on rotation of the drum, results in a substantial amount of ion channeling in the wafers being implanted. Ion channeling is a known problem where ions implanted in alignment with planes or axes of the crystal lattice of the substrate are channeled by these axes and planes to greater depths in the substrate. As a result, depth control of the implant process is prejudiced and a substantial number of implanted ions penetrate to excessive depths.
This is a significant problem for the process of exfoliation of thin wafer laminae following implantation by, for example, H+ ions.